Did quite some stuff recently.
1. HSPICE, ADS, ANsoft simulations.
2. MOSFET miller platform.
3. Filtering inductor design.
4. PCIe going thru.
5. loop stability revisit.
#1: control theory and loop stability.
a. system modeling, including Laplace equation writing out, transfer function simplification.
b. stability judgment.
- sys w/ roots only on the left half s plane are stable.
- the way to get an unstable system to stable is to cancel the roots on the right half s plane or on the Im axis.
- the char eq doesnt involve zero so I dont know how to judge the effect of zeros on s plane besides the pole cancellation function. Anyway the effect of zeros can be drawn out in Bode plot.
#2: The best way to understand a protocol is to play the analyzer.
a. by playing ana, you get the chance of repetitively visit the sequence of the protocol, together w/ various types of errors.
b. you can even think from the perspective of the ana designers on how to layout the display of a protocol info., and what combinations of trigger conditions are useful.
In one word, no play, no study :)